A Novel Mitragynine Analog with Low-Efficacy Mu Opioid Receptor Agonism Displays Antinociception with Attenuated Adverse Effects

Abstract

Mitragynine and 7-hydroxymitragynine (7OH) are the major alkaloids mediating the biological actions of the psychoactive plant kratom. To investigate the structure-activity relationships of mitragynine/7OH templates, we diversified the aromatic ring of the indole at the C9, C10, and C12 positions and investigated their G-protein and arrestin signaling mediated by mu opioid receptors (MOR). Three synthesized lead C9 analogs replacing the 9-OCH₃ group with phenyl (4), methyl (5), or 3'-furanyl [6 (SC13)] substituents demonstrated partial agonism with a lower efficacy than DAMGO or morphine in heterologous G-protein assays and synaptic physiology. In assays limiting MOR reserve, the G-protein efficacy of all three was comparable to buprenorphine. 6 (SC13) showed MOR-dependent analgesia with potency similar to morphine without respiratory depression, hyperlocomotion, constipation, or place conditioning in mice. These results suggest the possibility of activating MOR minimally (G-protein E_max ≈ 10%) in cell lines while yet attaining maximal antinociception in vivo with reduced opioid liabilities.

Figure 1.

(A) Structure of selected natural and semi-synthetic analogs. (B) Semi-synthesis of C9 mitragynine and 7OH derivatives. (C) Semi-synthesis of C10 mitragynine and 7OH derivatives. (D) Semi-synthesis of C12 mitragynine and 7OH derivatives. Reagents and conditions: (a) AlCl₃, EtSH, DCM, 0 °C, 5 h; (b) PhNTf₂, Et₃N, DCM, rt, 12 h; (c, yielding 9) phenylboronic acid, Pd(PPh₃)₄, K₂CO₃, MeOH, toluene, 80 °C, 8 h; (d, yielding 8) 3-furanylboronic acid, Pd(PPh₃)₄, K₂CO₃, MeOH, toluene, 80 °C, 8 h; (e, yielding 10) DABAL-Me₃, Pd₂(dba)₃, XPhos, THF, 60 °C, 8 h; (f) yielding 4, 5, and 6 (SC13) oxone, NaHCO₃, H₂O, acetone, 0 °C, 1 h. (g) ethylene glycol, PIFA, CH₃CN, 0 °C, 1 h; (h) NBS, DMF, 5 h, rt; (i) NaBH₃CN, AcOH, MeOH, reflux, 12 h; (j, yielding 13) phenylboronic acid, Pd(dppf)Cl₂, KOAc, THF, 70 °C, 6 h; (k, yielding 14) 3-furanylboronic acid, Pd(dppf)Cl₂, KOAc, THF, 70 °C, 6 h; (l, yielding 15) DABAL-Me₃, Pd₂(dba)₃, XPhos, THF, 60 °C, 8 h; (m, yielding 16, 17, and 18) oxone, NaHCO₃, H₂O, acetone, 0 °C, 1 h. (n) NBS, AcOH, 4 h, rt; (o, yielding 20) phenylboronic acid, Pd(PPh₃)₄, K₂CO₃, MeOH, toluene, 80 °C, 8 h; (p, yielding 21) 3-furanylboronic acid, Pd(PPh₃)₄, K₂CO₃, MeOH, toluene, 80 °C, 8 h; (q, yielding 22) DABAL-Me₃, Pd₂(dba)₃, XPhos, THF, 60 °C, 8 h; (r, yielding 23, 24, and 25) oxone, NaHCO₃, H₂O, acetone, 0 °C, 1 h.

Figure 2.

G-protein signaling, arrestin signaling, whole cell electrophysiology in rat VTA, and Gα-subtype screening of 4, 5, and 6 (SC13) and MOR controls in hMOR 4, 5, and 6 (SC13) are MOR partial agonists in cell-based assays, G-protein signaling assays, and in ephys assays. (A) Compounds 4, 5, and 6 (SC13) are MOR partial agonists with lower efficacy than morphine, fentanyl, and DAMGO in Gi-1 BRET assays. (B) 4, 5, and 6 (SC13) showed no measurable β-arrestin2 recruitment (<10%) in BRET arrestin recruitment assays compared to fentanyl and DAMGO in this assay. (C) In Nb33 recruitment assays measured using BRET assays in hMOR, 6 (SC13) had lower efficacy than DAMGO and morphine and similar efficacy to buprenorphine. (D) In Nb33 recruitment assays measured using BRET assays in mMOR, 6 (SC13) had lower efficacy than DAMGO and morphine and similar efficacy to buprenorphine. (E) Summary inhibition of electrically evoked IPSCs in VTA neurons in response to 5 μM DAMGO, 10 μM morphine, 10 μM 4, 10 μM 5, and 10 μM 6 (SC13), where each circle is one neuron. Horizontal bars indicate means. 4, 5, and 6 (SC13) show lower efficacy than DAMGO. (F) Mean time course of responses during bath application of 6 (SC13), n = 8 in whole cell electrophysiology in rat VTA. See Table S3 in the SI for values for panels (A–D). (G) TRUPATH heatmaps demonstrate how a panel of 7OH analogs, 4, 5, and 6 (SC13), and MOR agonists engage Gαi/o-class transducers with varying potency (G) and efficacy (H). Most ligands exhibit enhanced (GαZ) relative to other G-protein transducers. Heatmap colors represent mean log(EC₅₀) and normalized efficacy values; NR, no response, presented as a white square. Mean values and standard error are reported in the Supporting Information, Table S4. Data for all functional assays that were carried out in hMOR were normalized to E_max of DAMGO. The dose response curves were fit using a three-parameter logistic equation in GraphPad Prism, and the data are presented as mean EC₅₀(pEC₅₀ ± SEM) for assays run in triplicate.

Figure 3.

Binding modes and interactions of 4, 5, and 6 (SC13) compared to morphine. (A–C) Representative conformations of the most populated clusters from MD simulations of MOR bound to 4 (blue), 5 (teal), and 6 (SC13) (purple) (panels (A–C), respectively), compared to a representative conformation of MOR bound to morphine (orange). The crystal structure of active MOR corresponding to PDB ID: 5C1M was used as a starting point for all molecular docking and simulation studies. The protein is represented as a gray cartoon in the morphine–MOR complex. Residues identified in the best eight performing models on experimental data are indicated with sticks. Transmembrane helices 5 and 6 are not shown for clarity. (D) Differences (plot at the bottom) between average structural interaction fingerprints (SIFts) calculated for 4, 5, and 6 (SC13) (plot in the middle) and SIFts calculated for morphine (plot at the top).

Figure 4.

Compound 6 (SC13) shows MOR-dependent antinociception and lacks abuse potential, constipation, respiratory depression, and hyperlocomotion at equianalgesic morphine doses. (A) Antinociception time course. Groups of C57BL/6 J mice were subcutaneously (sc) administered 6 (SC13) and antinociception measured using the 55 °C tail withdrawal assay. Data are shown as mean % antinociception ± SEM. (A) Effect of 6 (SC13) at doses of 1, 3, and 10 mg/kg (n = 8 each group) with repeated measures over time. 6 (SC13) showed potent dose-dependent antinociception. (B) 6 (SC13) antinociception in KO mice. Antinociception effect of 6 (SC13) (10 mg/kg, sc,) was evaluated in groups of (n = 8) in WT, MOR KO, KOR KO, and DOR KO mice. Antinociception of 6 (SC13) remained intact in DOR KO (p = 0.13) and KOR KO (p = 0.058) mice, while it was found attenuated in MOR KO. Results for 6 (SC13) were analyzed with one-way ANOVA followed by Tukey’s post-hoc test, F_3,28 = 24.07, p < 0.0001, ****p < 0.0001 relative to WT, ns = p > 0.05 relative to WT. Attenuation of 6 (SC13) antinociception in MOR KO was also significantly greater compared to DOR KO and KOR KO mice (p < 0.0001 each; Tukey’s post hoc test). All values are expressed as the mean ± SEM. (C) Conditioned place preference or aversion (CPP/CPA). Place conditioning evaluation of 6 (SC13), morphine, and U50,488H, in C57BL/6 J mice after IP or sc administration. Following the determination of initial preconditioning preferences, mice were place-conditioned daily for 2 days with 6 (SC13) (15 mg/kg, sc; n = 23), U50,488H (30 mg/kg, IP; n = 28) or morphine (10 mg/kg, IP; n = 18). Mean differences in time spent on the drug-paired side ± SEM are presented. 6 (SC13) does not display significant CPP or CPA compared to the matching preconditioning preference (p < 0.05), as determined by unpaired t test with Welch’s correction. Morphine showed CPP (*p = 0.0140), and U50,488H showed CPA (****p < 0.0001) and were significantly different from matching preconditioning preference. (D) 6 (SC13) effects on gastrointestinal transit. Mice were administered morphine (10 mg/kg, sc) or 6 (SC13) (15 mg/kg, sc) or saline (0.9%, po) and then fed a charcoal meal. After 3 h, morphine significantly reduced the distance traveled by the charcoal through the intestines, consistent with the action of a MOR agonist 5.07 ± 0.57 cm, compared to 29.5 ± 1 cm for saline-treated mice; F_2,21 = 81.88, p < 0.0001; one-way ANOVA with Dunnett’s multiple-comparison test. In contrast, compound 6 (SC13) was without significant effect (30.8 ± 2.52 cm). (E) Respiratory rate. Mice were administered either vehicle (n = 12), morphine (30 mg/kg, sc; n = 12), or 6 (SC13) (45 mg/kg, sc; n = 12), and the breath rates was measured every 20 min for 180 min. Morphine administered sc caused reduction in the breath rate with respect to saline at 20 min (**p = 0.0021), 40 min (***p = 0.0003) and 60 min (**p = 0.0010) post drug administration. 6 (SC13) (45 mg/kg, sc) was not significantly different from vehicle control except at 180 (****p < 0.0001) and 200 min (*p = 0.0410) where it showed an increase in breath rates as determined by 2-way ANOVA followed by Dunnett’s multiple-comparison test. (F) Locomotor effect. Mice were administered either saline (n = 20), vehicle (n = 24), morphine (10 and 30 mg/kg, sc; n = 12 each), and 6 (SC13) (45 mg/kg, sc; n = 12), and the distance traveled by each group of mice was measured. No significant locomotor effects were observed with 6 (SC13) as determined by two-way ANOVA followed by Dunnett’s multiple-comparison test in comparison to the vehicle, while morphine showed significant hyperlocomotion at every time point compared to saline (p < 0.0001).